Risk-Aware Motion Planning with Learned Trajectory Primitives and Probabilistic Safety Assessment

2026-07-29

AI Summary

This work addresses the challenge of balancing safety and efficiency in urban autonomous driving by proposing a motion planning approach that integrates learning with optimization. The method employs radial basis function networks to generate trajectory primitives with low jerk and embeds them within a model predictive control–based optimization framework. An analytical probabilistic collision assessment mechanism is introduced to enable efficient and interpretable risk-aware trajectory selection. By preserving dynamic consistency and satisfying system constraints, the approach substantially reduces computational complexity. Experimental results across diverse urban scenarios demonstrate that, compared to baseline methods, the proposed solution significantly enhances risk awareness while markedly decreasing violations of vehicle kinematic and dynamic limits.

Abstract

This paper presents a radial basis function network (RBFN)-informed motion planning framework for safe and efficient urban autonomous driving. The proposed approach combines RBFN-based candidate trajectory generation with an analytic collision probability assessment and optimization-based trajectory refinement. The network learns jerk-minimal trajectories, enabling the MPC to operate within a reduced and dynamically consistent search space. Candidate motion primitives are selected based on an accurate probabilistic risk measure. This design decreases solver complexity while preserving safety and constraint satisfaction. The framework is evaluated in numerous urban driving scenarios. Results demonstrate improved risk awareness and fewer vehicle-limit violations compared to benchmark methods. The proposed approach integrates learning-based trajectories into optimization-based motion planning, thereby ensuring safety and interpretability.

Problem

• risk-aware motion planning
• autonomous driving
• probabilistic safety assessment
• trajectory primitives
• urban driving scenarios

Innovation

• risk-aware motion planning
• trajectory primitives
• probabilistic safety assessment
• radial basis function network
• model predictive control